How to Write a Methodology Section That Survives Peer Review

Reviewers reject what they cannot evaluate. A strong methodology section removes every reason for doubt by answering eight questions before the reviewer can ask them.

What reviewers are actually checking

A reviewer reading your methods section is asking one question on every line: could I reproduce this and would I trust the result? Everything below exists to answer yes. Address these eight components, in roughly this order, and you remove the most common grounds for a "major revisions" decision.

1. Research design and its justification

State the design — cross-sectional survey, time-lagged, experimental, multi-source — and, crucially, why it fits your question. If you claim mediation, a single-wave cross-sectional design is a weakness you must acknowledge. Naming the design and defending it pre-empts the reviewer's first objection.

2. Sample and statistical power

Describe the population, sampling frame, and recruitment. Then justify your sample size with an a priori power analysis, not a post-hoc apology. A correlation needs roughly 84 respondents to detect a medium effect at 80% power; mediation 150–250; complex SEM often 300+. Reporting "we collected what we could" invites rejection; reporting "we targeted N = 200 to detect a medium indirect effect at 80% power" signals competence.

3. Procedure

Walk through what a respondent actually experienced: how they were invited, what they consented to, the order of measures, any time lag, and any attention checks. Enough detail that a stranger could rebuild your study.

4. Measures

For each construct, report the source scale, number of items, response format, a sample item, and the reliability you obtained (Cronbach's α and, for SEM, composite reliability and AVE). Use validated scales wherever possible — reviewers distrust home-made measures. If you adapted a scale, say exactly how and why.

5. Common method bias remedies

If your predictor and outcome come from the same respondent at the same time, common method bias (CMB) is a live threat and reviewers will raise it. Show both procedural remedies (temporal or source separation, varied scale anchors, guaranteed anonymity) and statistical checks (Harman's single-factor test is the minimum; a common latent factor or marker variable is stronger). Addressing CMB before the reviewer does is one of the highest-return paragraphs you can write.

6. Analysis plan

State the analytic approach and why it matches the model: regression, PROCESS macro for mediation/moderation, or structural equation modelling for latent constructs and complex paths. Specify how you will test indirect effects (bootstrapped confidence intervals, not just Sobel) and what fit indices you will report.

7. Ethics

Name the approving body and approval number, confirm informed consent, and describe how data were anonymised and stored. Journals increasingly desk-reject papers without a clear ethics statement.

8. Transparency and open science

State whether the study was pre-registered, whether data and materials are available, and disclose any use of AI tools. Even a sentence ("materials available on request; analysis was not pre-registered, so confirmatory claims should be read as exploratory") builds credibility through honesty.

The three most common reasons methods get rejected

• Unjustified sample size. No power analysis, or a sample too small for the model.
• Unaddressed common method bias in a single-source cross-sectional design.
• Mismatch between design and claims — causal language on correlational data.

A copy-ready section order

Design → Participants and procedure → Sample size and power → Measures (one paragraph per construct) → Common method bias → Analytic strategy → Ethics → Transparency. Follow that order and your reader never has to hunt for an answer.

Final thought

A methods section is not where you show off; it is where you earn trust. Write it so that a sceptical reviewer runs out of objections before running out of paragraphs.